Luminescence Applications - ACS Publications - American Chemical

Chapter 1

Downloaded by GLASGOW UNIV on June 14, 2013 | http://pubs.acs.org Publication Date: December 30, 1989 | doi: 10.1021/bk-1989-0383.ch001

The Science of Luminescence Marvin C. Goldberg and Eugene R. Weiner U.S. Geological Survey, Box 25046, MS 424, Lakewood, CO 80225 The use of luminescence methods to measure the properties of chemical and biochemical systems has expanded so greatly in recent years that technical reviews and bibliographies no longer attempt to encompass the entire field, dealing instead with some restricted portion of the major subdivisions that have evolved. In the 1986 Fundamental Reviews of Analytical Chemistry, Wehry (1) lists 639 references in a biannual review titled "Molecular Fluorescence, Phosphorescence, and Chemiluminescence" which is focused mainly on analytical applications, and totally excludes the major topics of atomic fluorescence; molecular luminescence in flames, plasmas, and discharges; infrared fluorescence; solid state phosphor and semiconductor luminescence; radioluminescence; liquid scintillation counting; X-ray induced luminescence; and luminescence probing of biochemical, macromolecular, and micellar systems. Activity in each of the excluded areas is at least comparable to that in analytical applications of luminescence and the literature dealing with applications to biochemical, macromolecular, and micellar systems is by far the largest. Luminescence measurements have become an essential tool in nearly every aspect of molecular research. In contrast, the comparable biannual review in the 1956 Analytical Reviews of Analytical Chemistry (2) is titled simply "Fluorometric Analysis" and contains just 93 references, with no apologies for omission of any important topics. The immense growth in the luminescence literature during the period between these two reviews had little to do with developments in fundamental theory. It was mainly due to the availability of new instrumentation, such as the photomultiplier (around 1950), the laser (around 1960), transistor and microcircuit electronics (around 1970), and ready access to laboratory computers (around 1975). All aspects of luminescence theory now being used to interpret luminescence measurements have been known since the early 1900's and nearly all of the types of measurements now being made had been initiated with cruder techniques by 1930. We discuss here many of the latest techniques in luminescence analysis with selected highlights from the historical development of luminescence and a look at several recent developments in luminescence applications that appear likely to be important to future research. A Short History of Molecular

Luminescence

The Early Period. Luminescent phenomena such as the aurora borealis, phosphorescence of the sea, luminous animals and insects, and phosphorescent wood were the earliest of spectral observations because they require nothing more than the This chapter not subject to U.S. copyright Published 1989 American Chemical Society

In Luminescence Applications; Goldberg, M.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.


2

LUMINESCENCE APPLICATIONS

unaided eye as a detector. Harvey (3) reports that one of the earliest written references to luminescence appeared in the Chinese literature of around 1500-1000 B.C., describing glowworms and fireflies. Harvey gives a fascinating account of the many early attempts to explain luminescence, which took a definitely modern scientific approach by the seventeenth century, after experimental research had become a respectable activity for philosophers. Francis Bacon, the great advocate of observation over speculation, already had noted by 1605 that different kinds of luminescence seemed to have different origins, writing (3), "...it is not the property of fire alone to give light", and pointing out that "...sugar shineth only while i t is i n scraping; and salt water while it is i n dashing; glowworms have their shining while they live, or a little after; only scales of fish putrefied seem to be of the same nature with shining wood: and it is true, that all putrefaction hath with it an inward motion, as well as fire or light". This last comment has in it the seeds of the much later description of light and heat as generated by internal vibrations and propagated according to wave theory. The Bolognian stone, the first artificial inorganic phosphor, was discovered accidentally i n 1603 by Vincenzo Cascariolo, an Italian bootmaker and amateur alchemist (3). He had found some heavy stones near his home and tried to extract precious metals from them. After being heated i n his furnace, the stones had the ability to emit a reddish light for a lengthy time after being placed i n the dark. We now know that the Bolognian stone was mainly barium sulfate with probable traces of bismuth or manganese, and was converted to the sulfides by strong heating. This important discovery attracted the attention of Galileo and his contemporaries who recognized that the phosphor would not luminesce until after it had been exposed to natural light. Since light had to be absorbed into the stone, as a sponge absorbs water, before the stone would emit light, the behavior of the phosphor implied at the time that light was a material substance. This led Galileo to speculate that light must travel with a finite velocity, which he tried unsuccessfully to measure. The Italian mathematician Zucchi wrote in 1652 (3) that the Bolognian stone emitted more intensely if it was exposed to brighter light and that the color of the emitted light did not change when the stone was exposed to white light, or to light passed through green, yellow, or red colored glass (a primitive spectroscopic experiment). He concluded that the light is not simply absorbed and reemitted unchanged, as is water from a sponge, but, "rather it excites and unites with a spirituous substance contained i n the stone, and when the illumination has ceased, this substance gradually dissipates...". The first recorded observation of fluorescence from a solution (prompt emission, in contrast to the delayed emission called phosphorescence of the Bolognian stone, see Table I) appears to be by Nicolas Monardes (3) in 1565, who reported that when a cup made of a certain wood called lignum nephriticum was filled with water, a blue tinge could be observed. The luminescent properties of aqueous extracts of this wood intrigued many seventeenth century scientists. Athanasius Rircher, the first prominent scientist of this period to specialize in studies of luminescence (3), examined aqueous extracts of the wood in 1646, stating that the colors observed depended on the ambient light intensity in which the observations were made. Robert Boyle in 1664, Isaac Newton in 1672, and Robert Hooke in 1678 all discerned that Kircher was wrong and that the color of the wood extract depended on the angle of viewing, being yellow by transmitted light and blue by "reflected" light, meaning when viewed off the angle of transmission where a faint emission could be detected. Harvey (3) points out that although Kircher is generally regarded as the discoverer of fluorescence in solutions, Boyle was the first to describe some of the important chemical characteristics of fluorescing organic solutions. Boyle found that after many extractions, the wood lost its power to make a fluorescent solution and concluded that there must be an "essential salt" i n the wood that was responsible for the luminescence. He also reported that the light was generated without heat, that air (oxygen) was necessary and that extinguishing the light by evacuation was reversible, that the fluorescence was more visible i n dilute solutions, and that adding an acid destroyed the effect, which returned



1.

GOLDBERG AND WEINER

The Science of Luminescence

3

with the further addition of a base. A l l of these scientists believed they were observing a reflection phenomenon, not distinguishing clearly between emission and reflection. Frictional and impact heating of objects had led many scientists to the conclusion that heat was related to some sort of internal motion. Because light and heat often accompany one another, it was logical to look for similarities i n origin. Descartes, who advocated that matter was made up of small particles in constant motion and that fire was violent motion of these particles around their common center, believed that friction between the internal particles of matter generated the light. He developed an amazingly perceptive theory of light around 1637. Quoting Harvey (3), "...Descartes regarded light as rapid motion i n a subtle (elastic) fluid (the aether), made up of particles (the first element) invading all space, even the pores of solid bodies like glass. Its particles moved in vortices, not in straight lines. Space was a plenum, with no vacuum anywhere. Light was not a transmission of particles but was communicated by the push or pressure of one particle to another, as the motion of one end of a stick is felt at the other. The sun's particles were in agitation and transmitted this agitation instantaneously throughout the Universe." A similar theory was proposed around 1665 by Robert Hooke (3-4), who wrote that light consists of rapid vibrations propagated instantaneously, or very nearly so, over any distance and that hot bodies emit light because of internal vibrations. He argued that luminescence without heat also arose from the same kind of internal motions. By 1700, observations of all the various types of luminescence (Table I), except for crystalloluminescence, had been recorded and several were recognized as being of different origins. In 1718, Newton (6) wrote in favor of Hooke's hypothesis, that the incandescence of all luminous bodies, whether hot or cold, originated from "vibratory motions of their parts". Confirmation of this speculation had to await the developments of the spectroscope and quantum theory. Harvey (3) recounts how Newton's pinhole and prism arrangement were used by Zanotti (1748) and Dessaignes (1811) to study inorganic phosphors, by Melvill (1752) to examine the bright yellow sodium color in an alcohol flame containing sea salt, by Priestley (1767) to look at electroluminescence, and by Herschel (1822) for studying colored flames. None of these investigators were able to obtain an accurate spectrum with Newton's apparatus, although Herschel suggested that the dispersed light patterns might be useful for detection of small quantities of compounds. In 1802, Humphry Davy observed that this ultraviolet region could excite luminescence from a phosphor. A major instrumental improvement, necessary for the development of a practical spectroscope, was Wollaston's (7) introduction i n 1802 of a slit i n place of Newton's pinhole, with which he saw the dark lines in the solar spectrum later known as Frauenhofer lines. Wollaston did not recognize the significance of the dark lines and believed they simply separated the colors. These lines were rediscovered in 1814 by Frauenhofer (3-4) who had turned Wollaston's device into a practical spectroscope by adding a collimating tube and a telescopic eyepiece for viewing the image of the slit. Spectroscopic experiments by Bunsen and Kirchoff (8) in 1861 led to the interpretation of the Frauenhofer lines as being due to absorption of solar light by atoms and molecules i n the atmospheres of the earth and the sun. These studies laid the foundation for spectral analysis based on the fact that every gaseous chemical element has a unique line spectrum. The study of bioluminescence turned in the right direction in 1821 when Isaac François Macaire (3) suggested that the source of light i n the glowworm might be some organic compound rather than an inorganic phosphor, as was commonly assumed. Macaire had observed that all chemical reagents that cause albumin to coagulate also extinguish the glowworm's light and concluded that the luminous substance might be some form of albumin. Louis Pasteur (3) described in 1864 the spectrum of the light from the tropical luminous beetle Pyrophorous as continuous, without dark or light bands. In 1885, Raphael Dubois (3) obtained a cold water extract and a hot water extract from Pyrophorous which, when mixed together, reacted to produce light. Dubois named the



4


hot water extract luciferin and the cold water extract luciferase. He was able to demonstrate that luciferase and luciferin comprised an enzyme-substrate system. Dubois' work stimulated a resurgence of interest in bioluminescence. By 1889, M . W. Beijerinck (3) had used luminescent bacteria analytically to detect small amounts of oxygen. Stokes (10), who coined the word "fluorescence" from fluorspar, made a thorough scientific investigation of photoluminescence in 1852. He was the first to state clearly that fluorescence was an emission process. He showed that emission intensity depends on sample concentration and proposed the principle, now known as Stokes* law, which states that fluorescent emission is always of longer wavelength than the excitation. He reported that emission from certain crystals was polarized but emission from solutions was not polarized, whether or not the exciting light was polarized. He also described how a tube of quinine solution glowed brightly with a blue color when placed in the ultraviolet portion of the solar spectrum. Stokes commented on the high sensitivity of fluorescent emission and predicted that fluorescence analysis was "...likely to prove a great value in the separation of organic compounds". Edmund Bequerel (11) continued along the line of Stokes' studies, measuring the excitation and emission spectra of several phosphors, the influence of temperature, and many other parameters. Numerous scientists had observed that luminescence lifetimes appeared to vary considerably from compound to compound. Among the properties that Bequerel measured was the duration of luminescence after excitation ceased. For this purpose, he built the first phosphoroscope in 1858, with which he could 4

measure emission lifetimes as short as 10~ seconds. With his phosphoroscope, Bequerel identified two different types of decay kinetics, exponential and hyperbolic, which he attributed to monomolecular and bimolecular decay mechanisms. Bequerel's son Henri also was a distinguished luminescence investigator. Henri Bequerel has a special place in the history of science for accidentally discovering radioactivity in 1896 while studying the luminescence of some uranium salts (9). He found that the salts could expose photographic plates through their black paper wrapping. The first specific suggestion to use fluorescence in an analytical application appears to have been i n 1867 by Goppelsrôder (12), who proposed the analysis of Al(III) by the fluorescence of its morin chelate. Luminescence excited by cathode rays was described by Goldstein (13) in 1876 and by Crooks (14) in 1881. The term "luminescence" (the Greek translation of lucifer, meaning light bearer ) was introduced by Eilhardt Wiedemann (15) in 1888, in order to distinguish between light emission from thermally excited substances and light emission from molecules that have been excited by other means without increasing their average kinetic energy. Emission having an intensity that exceeds emission of the same frequency from a black body at the same temperature, was termed luminescence by Wiedemann. In other words, Wiedemann characterized luminescence by the fact that it does not obey Kirchoff s law of thermal emission and absorption from a black body. He also reported that the double salts of platinum emitted polarized light when excited by cathode rays, illustrating that luminescence can originate from different kinds of excitation. Wiedemann classified six different kinds of luminescence according to their method of excitation. Although excitation mechanisms are much better understood today, contemporary classifications of luminescence are essentially the same as Wiedemann's as is illustrated in Table I. The very first spectroscopic instruments, from Newton's prism and pinhole to Frauenhofer's simple spectroscope, were constructed to observe luminescence. Even though the great sensitivity of luminescence detection seemed to promise that luminescence would become an important tool for chemical analysis, the fact is that absorption spectroscopy was the first spectroscopic technique to be widely used. A t first glance, this may seem surprising since absorption spectroscopy is inherently less sensitive and had to await the development of more complex instrumentation, especially, electronically amplified detection.


1.

GOLDBERG AND WEINER


5

Table I. Types of Luminescence

Cause

Effect Fluorescence

Emission of light due to an "allowed" electronic transition between excited and ground states having the same spin multiplicity, usually singlet. Lifetimes for such transitions are typically around 10"® s. Originally it was believed that the onset of fluorescence was instantaneous (within 1 0 " to lQr? s) with the onset of radiation but the discovery of delayed fluorescence (16), which arises from thermal excitation from the lowest triplet state to the first excited singlet state and has a lifetime comparable to that for phosphorescence, makes this an invalid criterion. Specialized terms such as photoluminescence, cathodoluminescence, anodoluminescence, radioluminescence, and Xray fluorescence sometimes are used to indicate the type of exciting radiation. Emission of light due to a "forbidden" electronic transition between excited and ground states having different spin multiplicities, usually triplet to singlet. Lifetimes for such transitions typically range from milliseconds to minutes. Emission of light accompanying mild heating of substances and arising from thermally activated ion recombination. The sample has been previously excited by exposure to light or particle radiation and the energy "frozen in", to be released (not produced) by heating. It actually is a delayed phosphorescence. Emission of light accompanying the introduction of an electric field and arising from electronic excitation due to the presence of the electric field. Emission of light accompanying the passage of an electric current through aqueous solutions and arising from chemical reactions of chemiluminescent species produced during electrolysis. Emission of light accompanying the passage of intense sound waves through a liquid and arising from electrical discharges in the residual gas of cavities formed by the acoustic energy in the liquid. Emission of light accompanying the rubbing or crushing of certain crystals and arising from frictional charge separation at the crystal surfaces. Also called piezoluminescence. Emission of light accompanying the crystallization of certain crystals from solution and probably arising from cleavages occurring during the growth of individual crystals. Thus, it is a form of triboluminescence. Luminescence which appears when crystals dissolve is termed lyoluminescence. Emission of light from electronically excited species produced in a chemical reaction. Chemiluminescence from a biological system.


11

Phosphorescence

Thermoluminescence

Electroluminescence

Galvanoluminescence

Sonoluminescence

Triboluminescence

Crystalloluminescence

Chemiluminescence Bioluminescence



6


The high sensitivity of luminescence detection results from the fact that emitted photons are observed directly and many detectors, including the eye, will respond to small numbers of photons. Spectral absorption i n a dilute sample, on the other hand, which compares the light intensity transmitted through an absorbing sample with the intensity of light incident on the sample, requires the detection of a small difference between large signals and can only be observed with electronic instrumentation. Furthermore, many luminescent phenomena emit i n the high energy blue and ultraviolet regions of the spectrum, where numerous sensitive detectors are available, whereas absorption spectra of interest are frequently in the infrared spectral region where detectors are relatively inefficient. Nevertheless, even though absorption measurements are inherently less sensitive than luminescence measurements, the first extensively used analytical applications of molecular spectroscopy developed around absorption spectrometers. For an analytical technique to be widely useful, it must be selective as well as sensitive and this was the "Achilles heel" of luminescence spectroscopy. Room temperature luminescent emission is generally broad and featureless, attributes which, i n earlier times, usually negated its high sensitivity because band overlapping made multicomponent analyses impossible. The sensitivity of the method could not be fully utilized until its selectivity was improved. In fact, under conditions of low selectivity, the high sensitivity often was as much of a problem as a benefit because there were frequent interferences from trace impurities. Even though Jean Becquerel (17) discovered i n 1907 that cooling samples to liquid air temperatures narrowed the spectral features considerably and increased the information content of luminescent spectra, this approach was not practical for routine use and attracted little interest for chemical analysis. It was spectral absorption spectroscopy that became the earliest spectroscopic method widely used for molecular chemical analysis, and the prediction of Stokes (10) concerning the potential usefulness of fluorescence analysis was long delayed in its attainment. Even though absorption spectroscopy is less sensitive than luminescence spectroscopy it was found more useful because of its broader applicability. Many molecules that are not luminescent will absorb light and are subject to measurement whereas only those molecules that undergo a transition process when at an energy level above the ground state, will luminesce and are subject to luminescent measurements. The high degree of selectivity inherent i n luminescence analysis could not be realized until the fresh insights of quantum theory revealed new, potentially measurable luminescence parameters and stimulated the improvements i n instrumentation necessary to measure them. The Modern Period. Luminescence experiments after 1900, in what might be called the modern period of experimentation, were based on the ideas emerging from quantum mechanics. The development of quantum theory in the early 1900*8 gave a solid theoretical foundation to the vast accumulation of spectral data reported during the previous 200 years. Quantum theory led to a very complete understanding of luminescent emission, readily explaining prior observations and allowing predictions of new phenomena. As a result, luminescence was well understood theoretically by 1920-1930, at which time it was recognized that luminescence spectroscopy is inherently more selective than absorption spectroscopy because at least five independent luminescence variables can be measured which are characteristic of a given sample component: • emission intensity as a function of excitation wavelength. • excitation intensity as a function of emission wavelength. • excited state lifetime. emission polarization. • quantum yield. By comparison, the only variable measured i n absorption spectroscopy is transmission as a function of incident wavelength. Once quantum theory had revealed the potential 9



1.

GOLDBERG AND WEINER


7

of luminescence spectroscopy for studying molecular systems, the development of instruments for measuring the luminescence variables followed closely. It becomes difficult at this point to follow a central thread of historical development. As the insights of quantum theory continued to stimulate new approaches to all the sciences, discoveries and applications related to luminescence appeared rapidly i n many diverse fïelds. As a result, the remaining history necessarily becomes highly selective and many interesting and important aspects of recent luminescence history have surely been omitted. Resonance radiation was demonstrated by Robert Wood (18) in 1905 when he showed that when a bulb filled with sodium vapor at low pressure was irradiated with light from an intense sodium flame, the bulb emitted only the sodium D lines, with very narrow widths that corresponded to thermal Doppler broadening only, even though the D lines i n the higher temperature exciting source were much broader. By 1911, Selenyi (19) had shown experimentally that emission from fluorescein has a spatial intensity distribution that is characteristic of an electric dipole oscillator, confirming the theoretical explanation of the origin of polarized emission. Although Stokes (10) had not detected polarized fluorescence from solutions, Weigert (20) observed i n 1920, while studying fluorescent dye solutions, that such fluorescence is quite common. He found that the amount of polarization changes with solution viscosity and temperature, indicating that Brownian motion influences the polarization. Then between 1926-1929, Perrin (21-22) developed a theory for Weigert's Brownian depolarization and showed that differences i n the polarization of fluorescein in solutions of various viscosities were determined by the rotational rate and fluorescence lifetime of the molecules, a finding that foreshadowed today's extensive use of luminescence measurements to study the motions and the physical, chemical, and electronic environments of the emitting species. In 1922, Cario and Franck (23) demonstrated sensitized fluorescence in gaseous systems. A gas molecule that does not absorb energy from an illuminating light can be made to fluoresce by introducing another sensitizing molecule that does absorb from the light source. Cario and Franck illuminated a mixture of thallium and mercury vapors with the mercury resonance line, photoexciting only the mercury atoms. Emission of the characteristic thallium lines results when excited mercury atoms transfer their electronic energy to unexcited thallium atoms by collision. The first analysis of atomic fluorescence from metal atoms in flames was reported by Nichols and Howes (24) in 1924. Today, atomic fluorescence flame spectrometry is one of the most important techniques for elemental analysis of metals. Perrin observed in 1925 that increases in solution concentration were accompanied by increased depolarization and attributed this to energy transfer between like molecules (25), a far reaching concept that anticipated the discovery by Gaviola and Pringsheim (26) in 1927 that sensitized fluorescence occurs in liquid as well as i n gaseous systems. Gaviola and Pringsheim found, that electronic energy can be transferred over longer distances than exist in a colli si on al encounter, i n accordance with the quantum theory of nonradiative resonant electronic energy transfer (27). B y measuring the depolarization of fluorescein, immobilized i n a glycerol solution to eliminate the rotational depolarization observed by Perrin (21-22), they proved that depolarization also arises by electronic energy transfer from photoexcited molecules to unexcited molecules which then emit with a net loss of polarization, confirming Perrin's (25) conjecture about concentration depolarization. A n important result of their experiments was the finding that electronic energy can be transferred over distances nearly the order of the emission wavelength (50-100 nm). Then i n 1928, Warburg and Negelein (28) used the new technique of energy transfer depolarization to show that electronic energy could be transferred from the aromatic amino acids to the heme i n a heme protein. Perrin's treatment of energy transfer depolarization was later modernized by Fôrster (29) in 1948 i n terms of modern quantum theory, and Forster's treatment has been valuable for clarifying the behavior



8


of many photochemical processes where energy migration is important, as in photosynthesis (30). Methods based on Forster's theory have since developed into an elegant technique for using energy transfer phenomena as a molecular ruler, accurately measuring the distances between fluorophore positions, particularly in biochemical systems (31). Gaviola (32) reported the construction of the first phase fluorometer in 1927, with which he measured the fluorescence lifetime of aqueous rhodamine Β as 2 nanoseconds and that of fluorescein as 4.5 nanoseconds, in close agreement with the accepted values of today. In a phase fluorometer, the intensity of the excitation light source is sinusoidally modulated. The time lag between energy absorption and emission by sample molecules introduces a phase delay and demodulation, relative to the exciting waveform, into the emission from excited fluorophores. The phase delay and demodulation each can be used separately to calculate fluorescence lifetime. The resolution of Gaviola's device was about one nanosecond, a great improvement over Bequerel's (11) phosphoroscope. Teale (33) describes how visual detection was used for the phase measurement i n Gaviola's instrument; a polarizer was rotated to visually match a double image. The more obvious pulse excitation method for measuring fluorescence lifetimes, where the time decay of emission intensity is measured directly, was not to appear until around 1950, following the commercial availability of photomultipliers. Between 1923 and 1927, the concepts of quantum efficiency (number of photons emitted divided by number of photons absorbed by a sample) and quantum yield (fraction of excited molecules that emit) had been defined and values determined for many compounds by Vavilov (34). The quantum yield indicates the extent that other energy loss mechanisms compete with emission in an excited molecule. Although the quantum yield is influenced by the molecular environment of the emitter, for a given environment it depends on the nature of the emitting compound and is independent of concentration and excitation wavelength, at least at low concentrations (35). Thus, it serves as another measurable parameter that can be used to identify the compounds i n a sample and also, because of its sensitivity to the surroundings of the luminophore, to probe the environment of the emitter. The short period 19261927 was remarkably productive with regard to its impact on present day research. Techniques that are based on Gaviola's phase fluorometer, Perrin's theory of rotational depolarization, Vavilov's clarification of the concept of quantum yield, and the demonstration by Gaviola and Pringsheim of resonant energy transfer i n solutions, have become indispensable tools in all areas of molecular research. It is interesting to note that Gregorio Weber (36), in 1983, picked just these four discoveries as his choices for the four most important advances in the field of fluorescence spectroscopy of solutions and biochemical systems. In 1931, K. Weber (37) showed that the fluorescence spectrum of 1naphthylamine-4-sulfonate was affected differently by p H changes than was the absorption spectrum, illustrating that the different electron configurations of the ground and excited states may cause them to exhibit very different chemical behaviors. This finding initiated a continuing series of studies using luminescence to examine the unique chemistry of electronically excited species. Information about the chemical properties of excited species is essential to an understanding of chemical reactions initiated by the absorption of high energy quanta, as occur i n photochemistry and radiation chemistry. Fôrster and Kasper (38) in 1954 observed red-shifted emission from concentrated pyrene solutions that they attributed to the formation of a molecular dimer formed from two pyrene molecules, one excited and one in its ground state, later to be called an excimer by Stevens and Hutton (39). By 1955, Fôrster (40) and Weller (41) had measured excited state ionization constants of organic acids and determined that many have excited state values that differ by several orders of magnitude from the ground state values. In 1963 Leonhardt and Weller (42) detected emission from excited state complexes of unlike molecules, which later became known as exciplexes. Fôrster



1.

GOLDBERG AND WEINER


9

and Selinger (43) in 1964 studied the kinetics of excimer formation in micelles, which they were able to relate to diffusion rates and microviscosities within a micelle. Luminescence studies of excited state chemistry have evolved into a line of research that is helping to bring order into the study of excited state reactions by providing essential data for many species, such as excited state ionization constants (44), excited state proton transfer rates (45), kinetic mechanisms and rate constants for excimer and exciplex reactions (46), and relaxation characteristics of polar solvents (47). Jablonski (48-49) developed a theory in 1935 in which he presented the now standard "Jablonski diagram" of singlet and triplet state energy levels that is used to explain excitation and emission processes in luminescence. He also related the fluorescence lifetimes of the perpendicular and parallel polarization components of emission to the fluorophore emission lifetime and rate of rotation. In the same year, Szymanowski (50) measured apparent lifetimes for the perpendicular and parallel polarization components of fluorescein in viscous solutions with a phase fluorometer. It was shown later by Spencer and Weber (51) that phase shift methods do not give correct values for polarized lifetimes because the theory does not include the dependence on modulation frequency. In 1944, Lewis and Kasha (52) identified phosphorescence as a "forbidden" transition from an excited triplet state to the ground singlet state and suggested the use of phosphorescence spectra to identify molecules. Since then, phosphorimetry has developed into a popular method of analysis that, when compared with fluorometry, is more sensitive for some organic molecules and often provides complimentary information about structure, reactivity, and environmental conditions (53). Pulse fluorometry for directly measuring fluorescence and polarization decay times was developed during the 1950's, immediately following the commercial availability of photomultipliers (the R C A 1P28 photomultiplier introduced around 1950 quickly became ubiquitous in spectroscopy laboratories). Ware (54) placed in the introductory part of his paper on the current technique of pulse fluorometry, a brief outline of the early history of pulse methods, from the nanosecond measurements of Brody in 1957, who connected the output of his photomultiplier directly to the deflection plates of a Tektronix oscilloscope because pulse amplifiers at that time could not handle transient signals in the nanosecond range, to the picosecond measurements that first appeared in the 1970's using single-photon time-correlation methods with pulsed lasers and synchrotron radiation light sources. Single-photon counting time-correlation techniques evolved around 1960 (55-58) and has been crucial for the development of instruments that measure the time dependence of depolarization by pulse methods. Synchrotron radiation was first used to measure fluorescence lifetimes in 1973 (59). The cavity-dumped mode-locked C W laser was developed as a source for the single photon technique in the 1970's (60-61). The fluorescent lifetime of chlorophyll in vivo was first measured in 1957, independently by Brody and Rabinowitch (62) using pulse methods, and by Dmitrievskyand co-workers (63) using phase modulation methods. Because the measured quantum yield was lower than that predicted from the measured lifetime, it was concluded that much of the chlorophyll molecule was non-fluorescent, suggesting that energy transfer mechanisms were the means of moving absorbed energy to reactive parts of the molecule. Luminescence measurements on proteins occupy a large part of the biochemical literature. In what surely was one of the earliest scientific reports of protein photoluminescence uncomplicated by concurrent insect or microorganism luminescence, Beccari (64) , in 1746, detected a visible blue phosphorescence from chilled hands when they were brought into a dark room after exposure to sunlight. Stokes (10) remarked that the "dark" (ultraviolet) portion of the solar spectrum was most efficient in generating fluorescent emission and identified fluorescence from animal matter in 1852. In general, intrinsic protein fluorescence predominantly occurs between 300 nm and 400 nm and is very difficult to detect visually. The first



10


quantitative study of protein fluorescence, by Shore and Pardee (65) in 1956, had to await the development of high-pressure quartz xenon lamps, blazed diffraction gratings, large throughput spectrometers with quartz optics and sample holders, and UV-sensitive photomultipliers, all of which became available during the period 19501960. In 1952, Shpol'skii observed (66) that the fluorescence spectra of many aromatic molecules changed from broad bands to relatively sharp (~10 cm"*) fine structure when the aromatic molecules were dissolved in an n-alkane solvent and frozen to liquid nitrogen or liquid helium temperatures. The best results are obtained when the length of the solvent molecule closely matches the long axis length of the aromatic solute molecule. Under these conditions, the solute molecule can replace a solute molecule in the crystalline lattice without significant strain so that there is a reduction (relative to vitreous solvents) of the number of solvent cage orientations that are possible. This method of spectral line narrowing has become an important standard technique i n luminescence spectroscopy. Also i n 1952, Weber (67-68) introduced the idea of adding an extrinsic florescent probe to proteins, to facilitate measuring the steady state rotational depolarization of proteins and obtaining information about their size, conformation, and flexibility. With this work, Weber showed how the fluorescence polarization technique could be used to measure the Brownian motion of proteins. Proteins had been thought to adhere to a rather rigid "lock and key" mechanism in their reactions but the earliest fluorescence polarization measurements showed that many proteins were very flexible. The concept has since been highly developed by Weber and his co-workers (36, 51, 67-68) by combining steady state methods with phase modulation methods for measuring fluorescence lifetimes, depolarization, and differential phase polarization. A n indication of two or more lifetimes can arise when a single fluorophore is found in two or more very different surroundings, as long as the characteristics of the different surroundings remain distinct for a time of the order of the lifetime of the fluorescence. Much information can be learned about the in situ behavior of proteins by studying the causes for heterogeneity of lifetime measurements. Polarization and emission lifetime measurements of tryptophan or probe residues on proteins often show heterogeneous decay behavior that can be linked to dynamic exchange interactions between residues on different sites. Changes in quantum yields and emission maxima of intrinsic fluorescence also can be used to monitor conformational changes of proteins and their interaction with ligands (71). In 1947, William McElroy (72) discovered that adenosine triphosphate (ATP) and magnesium were essential for the bioluminescence reactions by which fireflies produce light. During the period 1956-69, McElroy and associates isolated the light producing substances luciferin and luciferase from fireflies and, along with other workers, described the reaction mechanism and kinetics (73). Because of the importance of A T P in the mechanism, firefly luciferase is used in a standard analytical procedure that is specific and extremely sensitive for A T P . The detection limit for the firefly assay method is about 1 0 ' mol/L A T P , allowing the detection of single cells in certain cases (74). Since A T P is present in all organisms with a fairly constant intercellular concentration, A T P concentration is a useful index of total biomass and has been applied as such to monitor bacteria in food, biomass i n environmental waters and soils, bacterial screening i n urine, and biomass i n wastewater and sewage sludge (75). The probable universality of A T P as a metabolite and the high sensitivity of its analysis suggest the use of the firefly assay technique as a tool for the detection of life on other planets. The use of phase sensitive detection with the phase fluorometer to analyze multicomponent systems was first described in 1970 by Veselova and coworkers (76). By using phase sensitive detection, the detector phase angle can be adjusted to be exactly out of phase with the phase-delayed emission from any single fluorophore, suppressing its contribution to the total emission signal. Phase sensitive detection, coupled with 1 1



1.

GOLDBERG AND WEINER


11

measurements at several different modulation frequencies can be used to separate the individual emission spectra and fluorescent lifetimes of the different components i n a mixture of fluorophores, providing the lifetimes are different enough to allow the required detector phase angles to be resolved. A n improved method using a crosscorrelation technique was been developed about the same time (77), in which the gain of the photomultiplier detector is modulated at a frequency slightly different from that of the excitation energy. This generates directly in the detector a low frequency crosscorrelation signal around 10 H z that carries the same phase and modulation information as the original high frequency. The phase and modulation of the low frequency cross-correlation signal can be measured with a better signal-to-noise ratio and permits much greater precision than direct measurements on the high frequency signal. A t present, lifetime differences around 1-10 ps are measurable with phase modulation and pulse methods. This allows a clear experimental separation of phenomena that occur in a time scale where molecular motions and interactions are important, around 1 ns to 1 μβ or longer, from those occurring within 0.1 ns to 1 ps, which arise from phenomena that reflect only internal energy changes unaffected by molecular motions. Weber (36) has referred to events occurring in the time domain shorter than 0.1 ns as belonging to the "private life" of molecules, while those occurring in a time interval around 1 ns or longer belong to their "public life". The story of luminescence advances since 1970 is best found in the current literature. It is too large and diversified to be readily summarized in a single article and too recent to edit for the most important events. Luminescence studies are so pervasive throughout all the fields of science that each special sphere of inquiry could have its own recent history of luminescence. Many of the most important developments of today will necessarily remain hidden to most scientists until later discoveries reveal their significance. For these reasons, this report continues with a personal selection of topics, chosen to guide the reader to a potpourri of interesting recent reports. A Look at Some Important Recent Developments in Luminescence The active state of luminescence spectrometry today may be judged by an examination of the 1988 issue of Fundamental Reviews of Analytical Chemistry (78), which divides its report titled "Molecular Fluorescence, Phosphorescence, and Chemiluminescence Spectrometry" into about 27 specialized topical areas, depending on how you choose to count all the subdivisions. This profusion of luminescence topics i n Fundamental Reviews is just the tip of the iceberg, because it omits all publications not primarily concerned with analytical applications. Fundamental Reviews does, however, represent a good cross-section of the available techniques because nearly every method for using luminescence in scientific studies eventually finds a use in some form of chemical analysis. Since it would be impossible to mention here all of the current important applications and developments in the entire universe of luminescence, this report continues with a look at progress in a few current areas that seem significant to the author for their potential impact on future work. Multidimensional measurements. A major use of luminescence measurements is to probe the environment of the emitting species. Many luminescence parameters are sensitive to the surroundings as well as the nature of the luminophore. The high sensitivity of luminescence measurements allows small changes i n measured parameters caused by environmental differences to be quantitatively measured. (See the paragraph below on lanthanide ion probes for an example.) Excitation and emission maxima, quantum yield, depolarization, temperature coefficient, and lifetimes all can be affected by environmental changes such as p H , ionic strength, temperature, viscosity, chemical binding, and unpaired electron spin density, making luminescence measurements inherently information-rich about the species emitting



12


and their environment. Luminescence is especially well suited for monitoring chemical processes that are sensitive to the immediate environment, as i n living cells. Multidimensional measurements combine two or more luminescence parameters, either simultaneously or i n sequence, into a single measurement in order to make possible a more complete utilization of the information present in the luminescent signal, permitting a more complete characterization of the luminescent material and increasing analytical selectivity. The combination of luminescence measurements with other techniques, such as chromatographic separation and chemical derivatization, also enhance analytical selectivity and could properly be included i n the category of multidimensional measurements. Miller (79) regards H P L C coupled with fluorescence detectors as one of the most important current applications of fluorescence. In chapter 12 of this book, Nelson et al, describe the development of derivitizing agents useful for making fluorophores from nonfluorescent compounds so they can be analyzed by H P L C with fluorescent detection. Even the fact that some species do not fluoresce may be used as a selectivity tool. Computers and modern electronics have made multidimensional measurements possible on a hitherto unprecedented scale and this approach has been one of the most effective ways to achieve high selectivity in luminescence measurements. Many ingenious techniques have recently become available for utilizing the multiple luminescence variables and these have made luminescence measurements a routine and valuable tool in almost all areas of experimental science. Many current multidimensional methods are based on instruments that combine measurements of several luminescence variables and present a multiparameter data set. The challenge of analyzing such complex data has stimulated the application of special mathematical methods (80-85) that are made practical only with the aid of computers. It is to be expected that future analytical strategies will rely heavily on computerized pattern recognition methods (79, 86) applied to libraries of standardized multidimensional spectra, a development that will require that published luminescence spectra be routinely corrected for instrumental artifacts. Warner et al, (84) have discussed the multiparameter nature of luminescence measurements in detail and list fourteen different parameters that can be combined i n various combinations for simultaneous measurement, thereby maximizing luminescence selectivity with multidimensional measurements. Table II is adapted from their paper with the inclusion of a few additional parameters. Several of the entries: excitation-emission matrix, synchronous spectra, and time -resolved depolarization, are themselves widely used multidimensional measurements that may be combined further with other parameters. Table II. Luminescence Parameters that can be Combined for Selectivity Enhancement Spectral measurements Excitation spectra Emission spectra

Lifetime jpeasurements Fluorescence Phosphorescence

Low-temperature (both time and phase spectra resolved) Excitation-emission matrix Synchronous spectra

Miscellaneous Polarization measurements _measurements Static depolarization Quenchometry Surfactant Differential phase enhancement (micelles, etc.) Depolarization Time-resolved depolarization

Quantum yield

Circular polarization

Two-photon excitation

Circular dichroism

Sensitized measurements Chemical derivatization Chromato graphic detection



1. GOLDBERG AND WEINER

13

The Science ofLuminescence

New ways to combine the various parameters of Table II are continually appearing in the literature. One of the most popular methods of multidimensional analysis is to measure the excitation-emission matrix (EEM), which allows plotting the emission intensity at all combinations of excitation and emission wavelengths in a single three-dimensional graph (81, 84-85, 87). Synchronous scanning (85, 88-89) is a simplification of the E E M technique in which both the excitation and emission wavelengths are scanned simultaneously with a typical offset of 50-100 nm. This approach generates a single two-dimensional slice through an E E M surface (90) and is especially useful for rapid screening analyses of complex mixtures. Synchronous scan luminescence spectrometry has recently been used i n combination with a supersonic jet expansion (91). By setting the wavelength difference to zero, a characteristic single very sharp emission line was obtained for each component i n a mixture of seven polyaromatic hydrocarbons. Paski and Blades (92) recently have found lifetime-wavelength resolved spectra useful for separating mixtures of inorganic powders which have been otherwise difficult to analyze. Lifetimes by the pulse method were determined as a function of emission wavelength and computer analysis of the data allowed separation of individual components from synthetic three-component mixtures. Wild (93) has compared E E M spectra of pyrene frozen i n a single crystal of η-heptane, i n a n-heptane Shporskii matrix, and i n ethanol glass to study the effects of different kinds of sites on the spectrum. E E M spectra are utilized in chapters 8 and 13 of this book. Hurtubise et al, show that an E E M spectra of a mixture of benzo(a)pyrene and benzo(e)pyrene adsorbed on an a-cyclodextrin-NaCl matrix allows the identification of nanogram levels of either compound i n the presence of an excess of the other. Theisen reports on the development of a catalog of E E M "signatures" for materials having luminescence stimulated by solar radiation, referenced to the solar Frauenhofer lines. Spectral line narrowing techniques. The usefulness of spectral measurements often depends on the wavelength resolution that can be attained. The broadness of spectral features i n many luminescence spectra limit the application of optical spectroscopy for fundamental measurements such as energy level structure, radiative decay behavior, and energy transfer mechanisms as well as for analytical identification. The main causes of line broadening are thermal Doppler broadening, instrumental distortions, and inhomogeneous broadening. The thermal broadening contribution to line width can be reduced by cryogenic cooling and supersonic jet cooling. Instrumental limitations on line sharpness are minimized by using the best available optical components and laser light sources. Inhomogeneous broadening arises from emitters that are in slightly different microenvironments, causing small differences in the energy levels of otherwise identical emitters. In solids, microenvironments are not uniform because of crystal strains and defects; in solutions, because of inhomogeneities of the solvent cages. After thermal broadening is minimized, the residual line widths are mainly due to inhomogeneous broadening and are still about 0.1-10 c m " i n many solid materials, and around 300 cm" for an organic compound in a vitreous solvent. Each emitter has a characteristic instrumentally limited homogeneous line width centered around the environmentally shifted energy level. The amount of improvement possible is indicated by the fact that the homogeneous line width at low temperatures (

Luminescence Applications - ACS Publications - American Chemical

Recommend Documents